The Indicator Service ships with a Docker Compose configuration that brings up the full stack: the API service, MongoDB, and Redis. A companion docker-compose.override.yml adds RabbitMQ and a development data producer.
The default RabbitMQ credentials (guest/guest) must be changed before deploying to any environment accessible over a network. Set RABBITMQ_DEFAULT_USER, RABBITMQ_DEFAULT_PASS, and update RABBITMQ_URL accordingly.
Services
The docker-compose.yml defines three services:
services:
indicator-service:
build: .
container_name: indicator-service
environment:
- ORIGINS=${ORIGINS:-localhost}
- REDIS_HOST=indicators-redis
- REDIS_PORT=6379
- MONGO_URI=mongodb://indicators-mongo/indicators
restart: on-failure
depends_on:
indicators-mongo:
condition: service_healthy
indicators-redis:
condition: service_started
networks:
- indicator-network
healthcheck:
test: ["CMD", "curl", "-f", "http://localhost:8080/health/"]
interval: 10s
timeout: 5s
retries: 3
start_period: 10s
develop:
watch:
- action: sync
path: ./app
target: /app
ignore:
- __pycache__/
- action: rebuild
path: requirements.txt
indicators-redis:
image: "redis:alpine"
container_name: indicators-redis
networks:
- indicator-network
restart: on-failure
indicators-mongo:
image: "mongo:latest"
container_name: indicators-mongo
environment:
MONGO_INITDB_DATABASE: indicators
volumes:
- indicators_db:/data/db
networks:
- indicator-network
healthcheck:
test: ["CMD", "mongosh", "--eval", "db.adminCommand('ping')"]
interval: 6s
timeout: 5s
retries: 5
start_period: 6s
restart: on-failure
networks:
indicator-network:
name: indicator-network
volumes:
indicators_db:
| Service | Image | Role |
|---|
indicator-service | Built from Dockerfile | FastAPI application |
indicators-mongo | mongo:latest | Primary data store |
indicators-redis | redis:alpine | Cache layer |
Health checks
The indicator-service polls http://localhost:8080/health/ every 10 seconds with a 5-second timeout and allows up to 3 retries before marking the container unhealthy. It also waits 10 seconds after startup before running the first check (start_period: 10s), giving uvicorn time to bind.
MongoDB is checked with mongosh --eval "db.adminCommand('ping')"every 6 seconds (5-second timeout, up to 5 retries). The indicator-service will not start until MongoDB passes this check.
Volumes and networking
indicators_db — a named Docker volume that persists MongoDB data across container restarts and recreations. Removing this volume will permanently delete stored indicator data.indicator-network — a bridge network shared by all services. Containers communicate using their service names as hostnames (e.g., indicators-mongo, indicators-redis).
Development override
docker-compose.override.yml is automatically merged when you run docker compose commands. It adds RabbitMQ and exposes the service port locally:
docker-compose.override.yml
services:
indicator-service:
environment:
- RABBITMQ_URL=amqp://guest:guest@indicators-rabbitmq/
depends_on:
rabbitmq:
condition: service_healthy
ports:
- "8080:8080"
rabbitmq:
image: "rabbitmq:3-management"
container_name: indicator-rabbitmq
ports:
- "5672:5672"
- "15674:15672"
environment:
- RABBITMQ_DEFAULT_USER=guest
- RABBITMQ_DEFAULT_PASS=guest
networks:
- indicator-network
healthcheck:
test: ["CMD", "rabbitmq-diagnostics", "check_port_connectivity"]
interval: 6s
timeout: 5s
retries: 5
start_period: 6s
restart: on-failure
data-producer:
build:
context: .
dockerfile: data-producer-example/data_producer/Dockerfile
container_name: indicator-data-producer
profiles: ["development"]
environment:
- RABBITMQ_URL=amqp://guest:guest@rabbitmq/
- API_URL=http://indicator-service:8080
depends_on:
rabbitmq:
condition: service_healthy
indicator-service:
condition: service_started
networks:
- indicator-network
restart: on-failure
data-producer service uses the development profile and is not started by default. Activate it with --profile development.
The RabbitMQ management UI is available at http://localhost:15674 when using the override file.
Deployment
Create a .env file
Copy .env.example and fill in your values:At minimum, set ORIGINS to your frontend’s origin and update the RabbitMQ credentials. Start the stack
Docker Compose waits for MongoDB to pass its health check before starting indicator-service. Verify the service is healthy
The indicator-service container should show healthy once the /health/ check passes. You can also query the endpoint directly:curl http://localhost:8080/health/
View logs
docker compose logs -f indicator-service
Hot reload development
Use docker compose watch to sync local code changes into the running container without a full rebuild:
The watch configuration in docker-compose.yml does the following:
| Trigger | Action |
|---|
Any file change under ./app/ | Synced directly into /app inside the container. uvicorn’s --reload flag picks up the change immediately. __pycache__/ is excluded. |
requirements.txt changes | Triggers a full image rebuild and container restart. |
Individual commands
# Stop and remove containers (data volume is preserved)
docker compose down
# Stop and remove containers AND the MongoDB data volume
docker compose down -v
# Rebuild only the service image
docker compose build indicator-service
# Restart only the service
docker compose restart indicator-service
# Run with the development data producer
docker compose --profile development up -d